Swapping internals in S-A hub?

[dabac]

Hi guys,

Are S-A hub shells identical across the range?

Aided by this forum I've been toying with various ideas on how to improve the usefulness of a composite wheel with a 3-speed hub.
And although I now have a a fairly good idea on how to get a slightly wider range and a few more intermediate gears by going to a triple front with carefully calculated tooth counts I still haven't got the range I'll need for a comfortable commute, using my current mtb-based ride as a reference.

So, is there some other option I should consider before I try to graft a Shimano-compatible DS hub flange onto the composite wheel by some cruel and unusual method?

For instance, would it be possible to transplant the internals from a S-A 5 speed into the hub shell of a 3-speed?

 

[Jenny Brien]

On Thu, 22 May 2008 12:34:12 +0100, dabac
<[email protected]> wrote:

>
> Hi guys,
>
> Are S-A hub shells identical across the range?
>
> Aided by this forum I've been toying with various ideas on how to
> improve the usefulness of a composite wheel with a 3-speed hub.
> And although I now have a a fairly good idea on how to get a slightly
> wider range and a few more intermediate gears by going to a triple front
> with carefully calculated tooth counts I still haven't got the range
> I'll need for a comfortable commute, using my current mtb-based ride as
> a reference.
>
> So, is there some other option I should consider before I try to graft
> a Shimano-compatible DS hub flange onto the composite wheel by some
> cruel and unusual method?
>
> For instance, would it be possible to transplant the internals from a
> S-A 5 speed into the hub shell of a 3-speed?
>
>
Yes, I've done it

 

[dabac]

On Thu, 22 May 2008 12:34:12 +0100, dabac
<[email protected]> wrote:

>
> Hi guys,
>
> Are S-A hub shells identical across the range?
>
> For instance, would it be possible to transplant the internals from a
> S-A 5 speed into the hub shell of a 3-speed?
>
>
Yes, I've done it

Good news! Now let's see what ebay has to offer...

 

[[email protected]]

someone wrote:

> Aided by this forum I've been toying with various ideas on how to
> improve the usefulness of a composite wheel with a 3-speed hub. And
> although I now have a a fairly good idea on how to get a slightly
> wider range and a few more intermediate gears by going to a triple
> front with carefully calculated tooth counts I still haven't got the
> range I'll need for a comfortable commute, using my current
> MTB-based ride as a reference.

> So, is there some other option I should consider before I try to
> graft a Shimano-compatible DS hub flange onto the composite wheel by
> some cruel and unusual method?

> For instance, would it be possible to transplant the internals from
> a S-A 5 speed into the hub shell of a 3-speed?

They are identical and the 5-speed is less likely to pop into neutral
when sprinting.

Jobst Brandt

 

[Tim McNamara]

In article <[email protected]>,
dabac <[email protected]> wrote:

> Hi guys,
>
> Are S-A hub shells identical across the range?
>
> Aided by this forum I've been toying with various ideas on how to
> improve the usefulness of a composite wheel with a 3-speed hub. And
> although I now have a a fairly good idea on how to get a slightly
> wider range and a few more intermediate gears by going to a triple
> front with carefully calculated tooth counts I still haven't got the
> range I'll need for a comfortable commute, using my current mtb-based
> ride as a reference.
>
> So, is there some other option I should consider before I try to
> graft a Shimano-compatible DS hub flange onto the composite wheel by
> some cruel and unusual method?
>
> For instance, would it be possible to transplant the internals from a
> S-A 5 speed into the hub shell of a 3-speed?

Yes, at least with the older made-in-Britain ones. Mark Stonich in
Minneapolis has done this mod several times. It's a straight swap.

 

[Frank Krygowski]

On May 22, 6:58 pm, Tim McNamara <[email protected]> wrote:
> In article <[email protected]>,
>
> > For instance, would it be possible to transplant the internals from a
> > S-A 5 speed into the hub shell of a 3-speed?
>
> Yes, at least with the older made-in-Britain ones. Mark Stonich in
> Minneapolis has done this mod several times. It's a straight swap.

I thought I read that the lowest gear of old S-A five speeds was
extremely inefficient. IOW, that shifting to low didn't really reduce
your workload significantly, but just gave you less speed.

I rode a friend's S-A five speed a few times, and got the impression
that the tale was true.

- Frank Krygowski

 

[[email protected]]

Frank Krygowski <[email protected]> wrote:

>>> For instance, would it be possible to transplant the internals
>>> from a S-A 5 speed into the hub shell of a 3-speed?

>> Yes, at least with the older made-in-Britain ones. Mark Stonich in
>> Minneapolis has done this mod several times. It's a straight swap.

> I thought I read that the lowest gear of old S-A five speeds was
> extremely inefficient. IOW, that shifting to low didn't really
> reduce your workload significantly, but just gave you less speed.

> I rode a friend's S-A five speed a few times, and got the impression
> that the tale was true.

So after riding up a few hundred yards of hill, was the hub noticeably
warm? My son's SA S5 has ordinary planetary gears that have no more
friction that other gears. Low gear is achieved through the stepped
planets that have a different number of teeth for input and output.
That's all that's different.

Jobst Brandt

 

[[email protected]]

On Thu, 22 May 2008 20:03:35 -0700 (PDT), Frank Krygowski
<[email protected]> wrote:

>I thought I read that the lowest gear of old S-A five speeds was
>extremely inefficient. IOW, that shifting to low didn't really reduce
>your workload significantly, but just gave you less speed.
>
>I rode a friend's S-A five speed a few times, and got the impression
>that the tale was true.
>
>- Frank Krygowski

Dear Frank,

Whatever the truth may be, this is a good example of the familiar RBT
controversy:

A poster heard about some effect, rode a bicycle, and things felt that
way--so was the effect real, or was it just expectation?

I couldn't find any efficiency testing of a 5-speed Sturmey-Archer,
but even a detailed test might not have settled the question.

After all, what if first gear was shown to be, say, 5.3% less
efficient than second gear? Is that increased work-load actually
noticeable while grinding up a hill at a 9.7% lower speed with [insert
imaginary number here] less wind drag?

The familiar Kyle & Berto efficiency study doesn't suggest that first
gear in hub drives is a hidden boat anchor:
http://www.ihpva.org/pubs/HP52.pdf

Sometimes that pdf downloads, sometimes it doesn't. It's worth saving
the damned file for later reference. Here's the table that compares
the efficiency, gear-by-gear, of some hub-drives at 80, 150, and 200
watts:

1 2 3 4 5 6 7 8 9 10 11 12 13 14
Maker/Speeds
Power Efficiency Percent**
Sachs 3 80 95.0 92.9 93.6
150 94.2 95.6 94.8
200 94.1 94.9 94.1
Shimano 3 80 90.5 93.5 87.2
150 93.0 93.9 88.6
200 93.2 95.0 87.2
Sturmey 3 80 92.3 95.4 91.8
150 93.3 95.3 91.8
200 93.0 95.6 91.8
Shimano 4 80 93.6 90.1 87.1 85.8
Automatic 150 95.6 90.9 88.9 87.0
200 95.3 92.8 90.0 88.0
Sachs 7 80 88.7 — — 89.2
150 89.9 — — 92.3
200* 91.0* — — 93.0*
Shimano 7 80 90.8 90.7 87.4 89.0 83.6 90.9 88.2
150 91.8 92.9 89.9 89.0 85.6 92.8 90.4
200 92.8 94.5 90.3 91.8 86.4 93.7 91.4
Sturmey 7 80 87.3 88.7 88.4 93.0 89.3 86.0 83.0
150 89.1 89.0 91.1 93.3 90.4 88.5 85.4
200 89.7 90.3 91.3 94.7 91.0 88.6 85.3
Rohloff 14 80 89.1 90.3 87.8 90.3 87.5 87.8 86.1 89.7 90.8 87.7 89.7
87.1 87.8 86.1
150 90.6 92.5 89.9 92.2 89.6 91.0 89.9 92.6 92.7 90.4 92.3
90.4 89.7 89.1
200 91.3 92.5 90.9 93.4 90.5 90.9 90.2 92.8 92.7 91.1 93.5
90.0 91.1 90.4

*The shift mechanism was broken, and would shift to only two gears.

** All efficiencies are uncorrected for the power consumed by the
ergometer wheel drive. Although this is not large, it would increase
the indicated efficiencies by 2 to 2.5% in most cases.

------------

Here's how 1st gear efficiency compares to 2nd gear from that table:

Sachs 3 +2.1% to -1.4%
Shimano 3 -0.3% to -0.9%
Shim auto 4 -2.5% to -5.7%
Sachs 7 (broken, no data)
Shimano 7 +0.1% to -1.7%
Sturmey 7 +0.1% to -1.4%
Rohloff 14 -1.2% to -1.9%

So a typical hub-drive first gear loses less than 2% of power compared
to second gear. That's less than the drag from a light generator.

Some riders might claim that they can detect such <2% losses (<4 watts
at 200 watts) even when they shift from one gear to another, but such
sensitivity would be impressive.

But Kyle & Berto didn't test the 5-speed Sturmey-Archer. Maybe that
hub has some design quirk that makes its first gear a real dog, unlike
the 3, 4, 7, and 14 speed hubs.

Cheers,

Carl Fogel

 

[A Muzi]

> Frank Krygowski <[email protected]> wrote:
>> I thought I read that the lowest gear of old S-A five speeds was
>> extremely inefficient. IOW, that shifting to low didn't really reduce
>> your workload significantly, but just gave you less speed.
>> I rode a friend's S-A five speed a few times, and got the impression
>> that the tale was true.

[email protected] wrote:
> Whatever the truth may be, this is a good example of the familiar RBT
> controversy:
>
> A poster heard about some effect, rode a bicycle, and things felt that
> way--so was the effect real, or was it just expectation?
>
> I couldn't find any efficiency testing of a 5-speed Sturmey-Archer,
> but even a detailed test might not have settled the question.
>
> After all, what if first gear was shown to be, say, 5.3% less
> efficient than second gear? Is that increased work-load actually
> noticeable while grinding up a hill at a 9.7% lower speed with [insert
> imaginary number here] less wind drag?
>
> The familiar Kyle & Berto efficiency study doesn't suggest that first
> gear in hub drives is a hidden boat anchor:
> http://www.ihpva.org/pubs/HP52.pdf
>
> Sometimes that pdf downloads, sometimes it doesn't. It's worth saving
> the damned file for later reference. Here's the table that compares
> the efficiency, gear-by-gear, of some hub-drives at 80, 150, and 200
> watts:
>
> 1 2 3 4 5 6 7 8 9 10 11 12 13 14
> Maker/Speeds
> Power Efficiency Percent**
> Sachs 3 80 95.0 92.9 93.6
> 150 94.2 95.6 94.8
> 200 94.1 94.9 94.1
> Shimano 3 80 90.5 93.5 87.2
> 150 93.0 93.9 88.6
> 200 93.2 95.0 87.2
> Sturmey 3 80 92.3 95.4 91.8
> 150 93.3 95.3 91.8
> 200 93.0 95.6 91.8
> Shimano 4 80 93.6 90.1 87.1 85.8
> Automatic 150 95.6 90.9 88.9 87.0
> 200 95.3 92.8 90.0 88.0
> Sachs 7 80 88.7 — — 89.2
> 150 89.9 — — 92.3
> 200* 91.0* — — 93.0*
> Shimano 7 80 90.8 90.7 87.4 89.0 83.6 90.9 88.2
> 150 91.8 92.9 89.9 89.0 85.6 92.8 90.4
> 200 92.8 94.5 90.3 91.8 86.4 93.7 91.4
> Sturmey 7 80 87.3 88.7 88.4 93.0 89.3 86.0 83.0
> 150 89.1 89.0 91.1 93.3 90.4 88.5 85.4
> 200 89.7 90.3 91.3 94.7 91.0 88.6 85.3
> Rohloff 14 80 89.1 90.3 87.8 90.3 87.5 87.8 86.1 89.7 90.8 87.7 89.7
> 87.1 87.8 86.1
> 150 90.6 92.5 89.9 92.2 89.6 91.0 89.9 92.6 92.7 90.4 92.3
> 90.4 89.7 89.1
> 200 91.3 92.5 90.9 93.4 90.5 90.9 90.2 92.8 92.7 91.1 93.5
> 90.0 91.1 90.4
>
> *The shift mechanism was broken, and would shift to only two gears.
>
> ** All efficiencies are uncorrected for the power consumed by the
> ergometer wheel drive. Although this is not large, it would increase
> the indicated efficiencies by 2 to 2.5% in most cases.
>
> ------------
>
> Here's how 1st gear efficiency compares to 2nd gear from that table:
>
> Sachs 3 +2.1% to -1.4%
> Shimano 3 -0.3% to -0.9%
> Shim auto 4 -2.5% to -5.7%
> Sachs 7 (broken, no data)
> Shimano 7 +0.1% to -1.7%
> Sturmey 7 +0.1% to -1.4%
> Rohloff 14 -1.2% to -1.9%
>
> So a typical hub-drive first gear loses less than 2% of power compared
> to second gear. That's less than the drag from a light generator.
>
> Some riders might claim that they can detect such <2% losses (<4 watts
> at 200 watts) even when they shift from one gear to another, but such
> sensitivity would be impressive.
>
> But Kyle & Berto didn't test the 5-speed Sturmey-Archer. Maybe that
> hub has some design quirk that makes its first gear a real dog, unlike
> the 3, 4, 7, and 14 speed hubs.

Efficiency, or lack thereof, is not at all obvious riding, at least to me.
(I switch semi-annually between an AW and a fixed gear. Same rims/tires.
Neither feels any more or less efficient. But then to Carl's comment,
my fixie has a tire driven dynamo. The other has a GH6 DynoHub.)

With Sturmey's S-5, gearing begins with AW's +33% high, direct, - 25%
low then adds an additional +10% 'super high' and -10% 'super low'. Ten
percent being a smallish difference, a rider who expects salvation with
a 'bail out' gear when clicking from 2 to 1 may be disappointed. And as
we know disappointment can lead to misdirected blame, "it's not efficient!".

--
Andrew Muzi
<www.yellowjersey.org/>
Open every day since 1 April, 1971
** Posted from http://www.teranews.com **

 

[jim beam]

[email protected] wrote:
> On Thu, 22 May 2008 20:03:35 -0700 (PDT), Frank Krygowski
> <[email protected]> wrote:
>
>> I thought I read that the lowest gear of old S-A five speeds was
>> extremely inefficient. IOW, that shifting to low didn't really reduce
>> your workload significantly, but just gave you less speed.
>>
>> I rode a friend's S-A five speed a few times, and got the impression
>> that the tale was true.
>>
>> - Frank Krygowski
>
> Dear Frank,
>
> Whatever the truth may be, this is a good example of the familiar RBT
> controversy:
>
> A poster heard about some effect, rode a bicycle, and things felt that
> way--so was the effect real, or was it just expectation?

nah, i noticed this stuff when i was a kid too small to sit the seat of
my mother's s.a. 3-speed bike. my derailleured dad's bike, i could ride
with one leg through the frame so i had comparison. nobody then was
telling me about planetary gear boxes. certainly not my mother.



>
> I couldn't find any efficiency testing of a 5-speed Sturmey-Archer,
> but even a detailed test might not have settled the question.
>
> After all, what if first gear was shown to be, say, 5.3% less
> efficient than second gear? Is that increased work-load actually
> noticeable while grinding up a hill at a 9.7% lower speed with [insert
> imaginary number here] less wind drag?
>
> The familiar Kyle & Berto efficiency study doesn't suggest that first
> gear in hub drives is a hidden boat anchor:
> http://www.ihpva.org/pubs/HP52.pdf
>
> Sometimes that pdf downloads, sometimes it doesn't. It's worth saving
> the damned file for later reference. Here's the table that compares
> the efficiency, gear-by-gear, of some hub-drives at 80, 150, and 200
> watts:
>
> 1 2 3 4 5 6 7 8 9 10 11 12 13 14
> Maker/Speeds
> Power Efficiency Percent**
> Sachs 3 80 95.0 92.9 93.6
> 150 94.2 95.6 94.8
> 200 94.1 94.9 94.1
> Shimano 3 80 90.5 93.5 87.2
> 150 93.0 93.9 88.6
> 200 93.2 95.0 87.2
> Sturmey 3 80 92.3 95.4 91.8
> 150 93.3 95.3 91.8
> 200 93.0 95.6 91.8
> Shimano 4 80 93.6 90.1 87.1 85.8
> Automatic 150 95.6 90.9 88.9 87.0
> 200 95.3 92.8 90.0 88.0
> Sachs 7 80 88.7 � � 89.2
> 150 89.9 � � 92.3
> 200* 91.0* � � 93.0*
> Shimano 7 80 90.8 90.7 87.4 89.0 83.6 90.9 88.2
> 150 91.8 92.9 89.9 89.0 85.6 92.8 90.4
> 200 92.8 94.5 90.3 91.8 86.4 93.7 91.4
> Sturmey 7 80 87.3 88.7 88.4 93.0 89.3 86.0 83.0
> 150 89.1 89.0 91.1 93.3 90.4 88.5 85.4
> 200 89.7 90.3 91.3 94.7 91.0 88.6 85.3
> Rohloff 14 80 89.1 90.3 87.8 90.3 87.5 87.8 86.1 89.7 90.8 87.7 89.7
> 87.1 87.8 86.1
> 150 90.6 92.5 89.9 92.2 89.6 91.0 89.9 92.6 92.7 90.4 92.3
> 90.4 89.7 89.1
> 200 91.3 92.5 90.9 93.4 90.5 90.9 90.2 92.8 92.7 91.1 93.5
> 90.0 91.1 90.4
>
> *The shift mechanism was broken, and would shift to only two gears.
>
> ** All efficiencies are uncorrected for the power consumed by the
> ergometer wheel drive. Although this is not large, it would increase
> the indicated efficiencies by 2 to 2.5% in most cases.
>
> ------------
>
> Here's how 1st gear efficiency compares to 2nd gear from that table:
>
> Sachs 3 +2.1% to -1.4%
> Shimano 3 -0.3% to -0.9%
> Shim auto 4 -2.5% to -5.7%
> Sachs 7 (broken, no data)
> Shimano 7 +0.1% to -1.7%
> Sturmey 7 +0.1% to -1.4%
> Rohloff 14 -1.2% to -1.9%
>
> So a typical hub-drive first gear loses less than 2% of power compared
> to second gear. That's less than the drag from a light generator.

it's not about gear to gear, it's about overall efficiencies in the
80's. that's not good. and below 200W, hub hears really do show
noticeably lower power than chain gears.


>
> Some riders might claim that they can detect such <2% losses (<4 watts
> at 200 watts) even when they shift from one gear to another, but such
> sensitivity would be impressive.

what's even more "impressive" is people standing in line to claim they
/can't/ tell when they haven't even done any physical comparison! could
they paradoxically be the same people claiming detectable difference in
tire rolling resistance?


>
> But Kyle & Berto didn't test the 5-speed Sturmey-Archer. Maybe that
> hub has some design quirk that makes its first gear a real dog, unlike
> the 3, 4, 7, and 14 speed hubs.
>
> Cheers,
>
> Carl Fogel

 

[[email protected]]

On Fri, 23 May 2008 12:35:36 -0700, jim beam
<[email protected]> wrote:

[snip]

>it's not about gear to gear, it's about overall efficiencies in the
>80's.

[snip]

Dear Jim,

Frank wrote about how dropping from second to first in a 5-speed
Sturmey-Archer seemed to reveal a dramatic difference in efficiency.

He might be right, he might have been expecting the difference after
hearing about it somewhere, or he might have been expecting a bigger
jump than the 10% change, as Andrew Muzi points out elsewhere in this
thread.

In any case, this sub-thread is specifically about gear-to-gear
efficiency differences in the same hub.

Cheers,

Carl Fogel

 

[Frank Krygowski]

On May 23, 4:34 pm, [email protected] wrote:
>
>
> Frank wrote about how dropping from second to first in a 5-speed
> Sturmey-Archer seemed to reveal a dramatic difference in efficiency.

Well, what I really felt was "Wow. That downshift didn't seem to help
much at all. I'm working just as hard even though I'm climbing
slower."

The situation: Bike was an ancient ladies frame Dunelt, owned by a
widow friend. It had been her husbands. (He'd had a much larger rear
cog brazed onto the original.)

We were visiting on vacation, and I was attempting to give the bike a
little tuneup for her. Their town, in the Appalachian foothills,
featured lots of short steep climbs (hence the big rear cog). She'd
warned me that her husband had some trouble keeping the hub gear
properly adjusted, and it demonstrated some trouble with a false
neutral. I fussed with cable adjustment and improved things, but
never got it to shift perfectly.

I've reassembled S-A three speeds, but never a five speed. I'm not
familiar enough with the guts to know if some bad component in there
could generate both shifting problems and extra inefficiency in low
gear.

This was long ago (probably 15 years), so my memory isn't to be
trusted; but IIRC, I read about the inefficient low _after_ my
experience with that bike, so it was a case of the reading confirming
my impression, not vice-versa.

- Frank Krygowski

 

[[email protected]]

On Fri, 23 May 2008 17:29:48 -0700 (PDT), Frank Krygowski
<[email protected]> wrote:

>On May 23, 4:34 pm, [email protected] wrote:
>>
>>
>> Frank wrote about how dropping from second to first in a 5-speed
>> Sturmey-Archer seemed to reveal a dramatic difference in efficiency.
>
>Well, what I really felt was "Wow. That downshift didn't seem to help
>much at all. I'm working just as hard even though I'm climbing
>slower."
>
>The situation: Bike was an ancient ladies frame Dunelt, owned by a
>widow friend. It had been her husbands. (He'd had a much larger rear
>cog brazed onto the original.)
>
>We were visiting on vacation, and I was attempting to give the bike a
>little tuneup for her. Their town, in the Appalachian foothills,
>featured lots of short steep climbs (hence the big rear cog). She'd
>warned me that her husband had some trouble keeping the hub gear
>properly adjusted, and it demonstrated some trouble with a false
>neutral. I fussed with cable adjustment and improved things, but
>never got it to shift perfectly.
>
>I've reassembled S-A three speeds, but never a five speed. I'm not
>familiar enough with the guts to know if some bad component in there
>could generate both shifting problems and extra inefficiency in low
>gear.
>
>This was long ago (probably 15 years), so my memory isn't to be
>trusted; but IIRC, I read about the inefficient low _after_ my
>experience with that bike, so it was a case of the reading confirming
>my impression, not vice-versa.
>
>- Frank Krygowski

Dear Frank,

I kept looking and found these extended comments, which support and
explain your impression of 1st gear in a 5-speed Sturmey-Archer being
less efficient than 2nd gear. The actual claim from S-A is a drop from
93% to 89%, but that might mean as much as 93.4% to 88.5%.

Just to confuse things, there were two versions of the 5-speed S-A
hub.

***

David Henshaw, Oct 98:

The 5-speed Sturmey is less efficient than the 3-speed in gears 1 and
5. Try riding up the same hill on a 3-speed in gear 1 and a 5-speed in
gear 1. The lower gearing on the 5 gives no real advantage. I was
sceptical until Andrew Ritchie, the man himself, demonstrated this.
Now I believe it. And the 5 is very adjustment-sensitive, unlike the
3, which lasts forever and ever.

On a 5-speed, I'd go for a 14-tooth sprocket, which gives about 8%
lower gearing. But, please remember to change the shim pack - the 14T
is thicker than the 13T... The 12% optional lower gearing is certainly
worth fitting on the 5-speed if you want really low gearing, but I'm
not sure about the 18% - a front changer is probably a more sensible
option.

I'm still happiest with a light, efficient 3-speed with standard
gearing. My clunky old bike still holds the unofficial Brompton speed
record, set at CycleFest 1996 by Richard Grigsby.

***

Willi Mindak, Nov 98:

I received an e- mail today from Marketing at SA. This was in reply to
an e- mail I send over a month ago. It says:

Dear Sir

The only information I have on this type relates to the 5 speed hub,
the information is as follows:-

Efficiency

Gear Ratios Efficiency
Super Low 0.667 89%
Low 0.789 93%
Normal 1.00 96%
High 1.266 94%
Super High 1.5 88%

Efficiency Test Conditions

Motor Torque Arm balance weight = 2.2Kg
Motor Speed = 64 RPM
No. of chainwheel teeth = 46
No. of Sprocket teeth = 18

I hope this information assists you.
Yours faithfully
Trevor Wilkinson
Marketing Executive

No surprises here. Legend had it that 1st and 5th gear were a bit
harder to work. Third gear is best with no internal ratio; I take it
the 4% are just bearing/ chain losses. I also asked for the 3 speed
values, but no joy. Anyway, these values are in the expected range.
They are actually better than I thought.

***

Why is highest and lowest gear in a 5 speed less efficient than
highest and lowest in a 3 speed?

Stein Somers, Oct 1998:

My guess: in the extremer gears, the planets in the cage roll on a
bigger central fixed cog, hence they are smaller in diameter and make
more rounds per minute. Both factors increase the friction at their
(simple) bearings. Maybe on top more than three planets are needed to
cope with the driving forces because the teeth on a smaller planet
make less firm contact with the central cog.

Come to think of it, the basic planetary system can only reach a 2:1
drive ratio using infinitely small planets. That's a 400% gear ratio,
low to high gear. (SA-3 has 178%, SA-5 has 225%). So how does the
illustrious Rohloff hub obtain more than 500%??

Carsten Thies, Jan 2003:

It has two identical planetary gears in line, giving 7 gears*, plus an
additional step-down planetary gear for the lower 7 gears.
* it's not 3x3=9 gears because step-up step-down or vice versa would
give the same as direct gear.
Custfold, Oct 1998:

The reason for the extreme high/extreme low being accessed by pulling
the second lever should give a clue - the old SA5 system used a second
cable to pull the dog over on the LHS of the unit and engage a step-up
ratio in the drive chain - otherwise the middle 3 were direct into the
gear cage. This in turn means a bigger slot in the axle and extra
little piece to slide in it for all 5 speed hubs. The Sprinter merely
uses a longer pull stroke to operate a 2-stage little bit in a slot
from 1 cable.

***

Everything above is from:
http://stein.dommel.be/brompton/chapters/Gears.html#HubGearEfficiency

Cheers,

Carl Fogel

 

[A Muzi]

>> [email protected] wrote:
>>> Frank wrote about how dropping from second to first in a 5-speed
>>> Sturmey-Archer seemed to reveal a dramatic difference in efficiency.

> Frank Krygowski <[email protected]> wrote:
>> Well, what I really felt was "Wow. That downshift didn't seem to help
>> much at all. I'm working just as hard even though I'm climbing
>> slower."
>>
>> The situation: Bike was an ancient ladies frame Dunelt, owned by a
>> widow friend. It had been her husbands. (He'd had a much larger rear
>> cog brazed onto the original.)
>>
>> We were visiting on vacation, and I was attempting to give the bike a
>> little tuneup for her. Their town, in the Appalachian foothills,
>> featured lots of short steep climbs (hence the big rear cog). She'd
>> warned me that her husband had some trouble keeping the hub gear
>> properly adjusted, and it demonstrated some trouble with a false
>> neutral. I fussed with cable adjustment and improved things, but
>> never got it to shift perfectly.
>>
>> I've reassembled S-A three speeds, but never a five speed. I'm not
>> familiar enough with the guts to know if some bad component in there
>> could generate both shifting problems and extra inefficiency in low
>> gear.
>>
>> This was long ago (probably 15 years), so my memory isn't to be
>> trusted; but IIRC, I read about the inefficient low _after_ my
>> experience with that bike, so it was a case of the reading confirming
>> my impression, not vice-versa.

[email protected] wrote:
> I kept looking and found these extended comments, which support and
> explain your impression of 1st gear in a 5-speed Sturmey-Archer being
> less efficient than 2nd gear. The actual claim from S-A is a drop from
> 93% to 89%, but that might mean as much as 93.4% to 88.5%.
>
> Just to confuse things, there were two versions of the 5-speed S-A
> hub.
>
> ***
>
> David Henshaw, Oct 98:
>
> The 5-speed Sturmey is less efficient than the 3-speed in gears 1 and
> 5. Try riding up the same hill on a 3-speed in gear 1 and a 5-speed in
> gear 1. The lower gearing on the 5 gives no real advantage. I was
> sceptical until Andrew Ritchie, the man himself, demonstrated this.
> Now I believe it. And the 5 is very adjustment-sensitive, unlike the
> 3, which lasts forever and ever.
>
> On a 5-speed, I'd go for a 14-tooth sprocket, which gives about 8%
> lower gearing. But, please remember to change the shim pack - the 14T
> is thicker than the 13T... The 12% optional lower gearing is certainly
> worth fitting on the 5-speed if you want really low gearing, but I'm
> not sure about the 18% - a front changer is probably a more sensible
> option.
>
> I'm still happiest with a light, efficient 3-speed with standard
> gearing. My clunky old bike still holds the unofficial Brompton speed
> record, set at CycleFest 1996 by Richard Grigsby.
>
> ***
>
> Willi Mindak, Nov 98:
>
> I received an e- mail today from Marketing at SA. This was in reply to
> an e- mail I send over a month ago. It says:
>
> Dear Sir
>
> The only information I have on this type relates to the 5 speed hub,
> the information is as follows:-
>
> Efficiency
>
> Gear Ratios Efficiency
> Super Low 0.667 89%
> Low 0.789 93%
> Normal 1.00 96%
> High 1.266 94%
> Super High 1.5 88%
>
> Efficiency Test Conditions
>
> Motor Torque Arm balance weight = 2.2Kg
> Motor Speed = 64 RPM
> No. of chainwheel teeth = 46
> No. of Sprocket teeth = 18
>
> I hope this information assists you.
> Yours faithfully
> Trevor Wilkinson
> Marketing Executive
>
> No surprises here. Legend had it that 1st and 5th gear were a bit
> harder to work. Third gear is best with no internal ratio; I take it
> the 4% are just bearing/ chain losses. I also asked for the 3 speed
> values, but no joy. Anyway, these values are in the expected range.
> They are actually better than I thought.
>
> ***
>
> Why is highest and lowest gear in a 5 speed less efficient than
> highest and lowest in a 3 speed?
>
> Stein Somers, Oct 1998:
>
> My guess: in the extremer gears, the planets in the cage roll on a
> bigger central fixed cog, hence they are smaller in diameter and make
> more rounds per minute. Both factors increase the friction at their
> (simple) bearings. Maybe on top more than three planets are needed to
> cope with the driving forces because the teeth on a smaller planet
> make less firm contact with the central cog.
>
> Come to think of it, the basic planetary system can only reach a 2:1
> drive ratio using infinitely small planets. That's a 400% gear ratio,
> low to high gear. (SA-3 has 178%, SA-5 has 225%). So how does the
> illustrious Rohloff hub obtain more than 500%??
>
> Carsten Thies, Jan 2003:
>
> It has two identical planetary gears in line, giving 7 gears*, plus an
> additional step-down planetary gear for the lower 7 gears.
> * it's not 3x3=9 gears because step-up step-down or vice versa would
> give the same as direct gear.
> Custfold, Oct 1998:
>
> The reason for the extreme high/extreme low being accessed by pulling
> the second lever should give a clue - the old SA5 system used a second
> cable to pull the dog over on the LHS of the unit and engage a step-up
> ratio in the drive chain - otherwise the middle 3 were direct into the
> gear cage. This in turn means a bigger slot in the axle and extra
> little piece to slide in it for all 5 speed hubs. The Sprinter merely
> uses a longer pull stroke to operate a 2-stage little bit in a slot
> from 1 cable.
>
> ***
>
> Everything above is from:
> http://stein.dommel.be/brompton/chapters/Gears.html#HubGearEfficiency

From Somers above "Maybe on top more than three planets are needed "

S5 and AW each have four planet gears (Brit: pinions), not three.

p.s. SW have three, but SW of course 'Seldom Work';AWs 'Always Work'!

p.p.s Yes I know it's 'Type A, Wide Range' but the mnemonic is cute.
--
Andrew Muzi
<www.yellowjersey.org/>
Open every day since 1 April, 1971
** Posted from http://www.teranews.com **

 

[James Thomson]

"A Muzi" <[email protected]> a écrit:

> With Sturmey's S-5, gearing begins with AW's +33% high,
> direct, - 25% low then adds an additional +10% 'super high'
> and -10% 'super low'. Ten percent being a smallish difference,
> a rider who expects salvation with a 'bail out' gear when clicking
> from 2 to 1 may be disappointed.

It's a bit more than 10% (the middle three gears are spaced a touch narrower
than an AW). The two ratios are 15:19 (19:15) and 2:3 (3:2). That gives (as
a percentage of direct drive):

66.7 78.9 100 126.6 150

James Thomson

 

[James Thomson]

"Frank Krygowski" <[email protected]> a écrit:

> I thought I read that the lowest gear of old S-A five speeds
> was extremely inefficient. IOW, that shifting to low didn't
> really reduce your workload significantly, but just gave you
> less speed.

That's an old myth. There's some relevant discussion in this thread:

http://groups.google.fr/group/rec.bicycles.tech/browse_thread/thread/38ba32282ea4ff3d/

James Thomson

 

[Tim McNamara]

In article <[email protected]>,
A Muzi <[email protected]> wrote:

> From Somers above "Maybe on top more than three planets are needed "
>
> S5 and AW each have four planet gears (Brit: pinions), not three.

My Sachs Torpedo 3 speed hub has, as far as I have found out from
searching on the net, 3 planet gears and that hub has a reputation for
reliability that is better than the AW (no walking out of engagement in
high gear). It also rated a bit higher in efficiency in Frank Berto's
tests- perhaps due to having three rather than four planet gears?

> p.s. SW have three, but SW of course 'Seldom Work';AWs 'Always Work'!
>
> p.p.s Yes I know it's 'Type A, Wide Range' but the mnemonic is cute.

Well, perhaps "'Twill Always Work Right" would fit.

 

[Frank Krygowski]

On May 24, 1:59 am, "James Thomson" <[email protected]> wrote:
> "Frank Krygowski" <[email protected]> a écrit:
>
> > I thought I read that the lowest gear of old S-A five speeds
> > was extremely inefficient. IOW, that shifting to low didn't
> > really reduce your workload significantly, but just gave you
> > less speed.
>
> That's an old myth. There's some relevant discussion in this thread:
>
> http://groups.google.fr/group/rec.bicycles.tech/browse_thread/thread/...
>
> James Thomson

Hmm. Not much meaningful discussion, though, from what I can see.
Just two people disagreeing, with no relevant data.

It would be nice to see some test data on the S-A 5 speed.

- Frank Krygowski

 

[Tim McNamara]

In article
<0ee921d9-7339-4279-b1af-dac25d4bc0d3@j22g2000hsf.googlegroups.com>,
Frank Krygowski <[email protected]> wrote:

> On May 24, 1:59 am, "James Thomson" <[email protected]> wrote:
> > "Frank Krygowski" <[email protected]> a écrit:
> >
> > > I thought I read that the lowest gear of old S-A five speeds was
> > > extremely inefficient. IOW, that shifting to low didn't really
> > > reduce your workload significantly, but just gave you less speed.
> >
> > That's an old myth. There's some relevant discussion in this
> > thread:
> >
> > http://groups.google.fr/group/rec.bicycles.tech/browse_thread/thread
> > /...
> >
> > James Thomson
>
> Hmm. Not much meaningful discussion, though, from what I can see.
> Just two people disagreeing, with no relevant data.
>
> It would be nice to see some test data on the S-A 5 speed.

In case your news server missed it, someone posted information they got
from S-A about the efficiencies of the S-A 5 speed in the various gears.
There was a pretty big drop in efficiency in gears 1 and 5. It should
just a few posts back up the thread.

 

[Dan Burkhart]

>> [email protected] wrote:
>>> Frank wrote about how dropping from second to first in a 5-speed
>>> Sturmey-Archer seemed to reveal a dramatic difference in efficiency.

> Frank Krygowski <[email protected]> wrote:
>> Well, what I really felt was "Wow. That downshift didn't seem to help
>> much at all. I'm working just as hard even though I'm climbing
>> slower."
>>
>> The situation: Bike was an ancient ladies frame Dunelt, owned by a
>> widow friend. It had been her husbands. (He'd had a much larger rear
>> cog brazed onto the original.)
>>
>> We were visiting on vacation, and I was attempting to give the bike a
>> little tuneup for her. Their town, in the Appalachian foothills,
>> featured lots of short steep climbs (hence the big rear cog). She'd
>> warned me that her husband had some trouble keeping the hub gear
>> properly adjusted, and it demonstrated some trouble with a false
>> neutral. I fussed with cable adjustment and improved things, but
>> never got it to shift perfectly.
>>
>> I've reassembled S-A three speeds, but never a five speed. I'm not
>> familiar enough with the guts to know if some bad component in there
>> could generate both shifting problems and extra inefficiency in low
>> gear.
>>
>> This was long ago (probably 15 years), so my memory isn't to be
>> trusted; but IIRC, I read about the inefficient low _after_ my
>> experience with that bike, so it was a case of the reading confirming
>> my impression, not vice-versa.

[email protected] wrote:
> I kept looking and found these extended comments, which support and
> explain your impression of 1st gear in a 5-speed Sturmey-Archer being
> less efficient than 2nd gear. The actual claim from S-A is a drop from
> 93% to 89%, but that might mean as much as 93.4% to 88.5%.
>
> Just to confuse things, there were two versions of the 5-speed S-A
> hub.
>
> ***
>
> David Henshaw, Oct 98:
>
> The 5-speed Sturmey is less efficient than the 3-speed in gears 1 and
> 5. Try riding up the same hill on a 3-speed in gear 1 and a 5-speed in
> gear 1. The lower gearing on the 5 gives no real advantage. I was
> sceptical until Andrew Ritchie, the man himself, demonstrated this.
> Now I believe it. And the 5 is very adjustment-sensitive, unlike the
> 3, which lasts forever and ever.
>
> On a 5-speed, I'd go for a 14-tooth sprocket, which gives about 8%
> lower gearing. But, please remember to change the shim pack - the 14T
> is thicker than the 13T... The 12% optional lower gearing is certainly
> worth fitting on the 5-speed if you want really low gearing, but I'm
> not sure about the 18% - a front changer is probably a more sensible
> option.
>
> I'm still happiest with a light, efficient 3-speed with standard
> gearing. My clunky old bike still holds the unofficial Brompton speed
> record, set at CycleFest 1996 by Richard Grigsby.
>
> ***
>
> Willi Mindak, Nov 98:
>
> I received an e- mail today from Marketing at SA. This was in reply to
> an e- mail I send over a month ago. It says:
>
> Dear Sir
>
> The only information I have on this type relates to the 5 speed hub,
> the information is as follows:-
>
> Efficiency
>
> Gear Ratios Efficiency
> Super Low 0.667 89%
> Low 0.789 93%
> Normal 1.00 96%
> High 1.266 94%
> Super High 1.5 88%
>
> Efficiency Test Conditions
>
> Motor Torque Arm balance weight = 2.2Kg
> Motor Speed = 64 RPM
> No. of chainwheel teeth = 46
> No. of Sprocket teeth = 18
>
> I hope this information assists you.
> Yours faithfully
> Trevor Wilkinson
> Marketing Executive
>
> No surprises here. Legend had it that 1st and 5th gear were a bit
> harder to work. Third gear is best with no internal ratio; I take it
> the 4% are just bearing/ chain losses. I also asked for the 3 speed
> values, but no joy. Anyway, these values are in the expected range.
> They are actually better than I thought.
>
> ***
>
> Why is highest and lowest gear in a 5 speed less efficient than
> highest and lowest in a 3 speed?
>
> Stein Somers, Oct 1998:
>
> My guess: in the extremer gears, the planets in the cage roll on a
> bigger central fixed cog, hence they are smaller in diameter and make
> more rounds per minute. Both factors increase the friction at their
> (simple) bearings. Maybe on top more than three planets are needed to
> cope with the driving forces because the teeth on a smaller planet
> make less firm contact with the central cog.
>
> Come to think of it, the basic planetary system can only reach a 2:1
> drive ratio using infinitely small planets. That's a 400% gear ratio,
> low to high gear. (SA-3 has 178%, SA-5 has 225%). So how does the
> illustrious Rohloff hub obtain more than 500%??
>
> Carsten Thies, Jan 2003:
>
> It has two identical planetary gears in line, giving 7 gears*, plus an
> additional step-down planetary gear for the lower 7 gears.
> * it's not 3x3=9 gears because step-up step-down or vice versa would
> give the same as direct gear.
> Custfold, Oct 1998:
>
> The reason for the extreme high/extreme low being accessed by pulling
> the second lever should give a clue - the old SA5 system used a second
> cable to pull the dog over on the LHS of the unit and engage a step-up
> ratio in the drive chain - otherwise the middle 3 were direct into the
> gear cage. This in turn means a bigger slot in the axle and extra
> little piece to slide in it for all 5 speed hubs. The Sprinter merely
> uses a longer pull stroke to operate a 2-stage little bit in a slot
> from 1 cable.
>
> ***
>
> Everything above is from:
> http://stein.dommel.be/brompton/chapters/Gears.html#HubGearEfficiency

From Somers above "Maybe on top more than three planets are needed "

S5 and AW each have four planet gears (Brit: pinions), not three.

p.s. SW have three, but SW of course 'Seldom Work';AWs 'Always Work'!

p.p.s Yes I know it's 'Type A, Wide Range' but the mnemonic is cute.
--
Andrew Muzi
<www.yellowjersey.org/>
Open every day since 1 April, 1971
** Posted from http://www.teranews.com **

Modern incarnations have triple planet sets, not quads. I'm looking at the internals of a current production 5 speed right now.
Dan Burkhart
www.boomerbicycle.ca